Arduino and ESP32-CAM Based Smart Waste Segregation System Using Sensors and Machine Learning

Improper segregation of waste at the source leads to inefficient recycling, environmental pollution, and health hazards. This paper presents an Arduino and ESP32-CAM based smart waste segregation system that combines sensor-based detection with machine learning-based image classification. The ESP32-CAM captures images of waste items, while a trained model identifies categories such as biodegradable, medical, plastic, metal, glass, and wood. Moisture, inductive, capacitive, and proximity sensors provide additional material information. Arduino processes the sensor readings and classification output and controls servo motors to direct waste into the appropriate bins. The proposed hybrid approach aims to reduce manual effort, improve sorting reliability, and support cleaner waste management. The system can be extended with IoT-based monitoring for use in homes, hospitals, institutions, and smart city environments.